The Cobblers Stick to Their Lasts: Pollinators Prefer Native Overalien Plant Species In

The cobblers stick to their lasts: Pollinators prefer native overalien plant species in a multi-species experiment

Biological Invasions

Thomas Chrobock*,#, Pius Winiger, Markus Fischer and Mark van Kleunen

*Corresponding author; Institute of Plant Sciences, University of Bern, Altenbergrain 21, CH – 3013 Bern, Switzerland, email: , telephone +41 (0)31 631 4938, telefax +41 (0)31 631 4942

#Present address: Carl Zeiss Naturzentrum Amrum, Strunwai 31, D – 25946 Norddorf auf Amrum, Germany

Online Resource

Online Resource Table 1 List of plant species used in our study with their breeding systems according to the BiolFlor database ( and the five pollinator groups that we observed most frequently on them. The pollinator groups are given in order of descending frequency. On five plant species, we observed fewer than five pollinator groups. The codes for the breeding systems are X (allogamous), XF (facultative allogamous), AFXF (mixed mating) and AF (facultative autogamous).

Online Resource Table 2 Numbers of plants of each study species that we used in each of the six study locations.

Online ResourceTable 3 Results of the statistical analyses of pollinator visitation per plant of native and related alien plant species (belonging to Asteraceae, Campanulaceae and Caryophyllaceae) in urbanized and semi-natural habitats in Switzerland. Subscript numbers denote error term used for calculating quasi-F- and p-values.

Error terms: 1: Species; 2: Location; 3: Species × Habitat type; 4: Species × Location; 5: Observation session; 6: Residuals. #: covariables had a positive effect on response variables. †: The proportion of variation explained by each factor can be estimated using the change in deviance (i.e. the mean deviance multiplied by the d.f.) relative to the total deviance; d.f.: degrees of freedom, ***p<0.001, **0.001<p<0.01, *0.01<p<0.05

Online Resource Figure 1Log10 (time a pollinator spent on a plant) observed within 30 minutes in urbanized and semi-natural habitats in Switzerland. Missing error bars indicate species with pollinator visitation in only one of the observation periods in the respective habitat type, shown are means ±1SE

Online Resource Table 4 Results of the statistical analysis of the time a pollinator spent on native and related alien plant species (Asteraceae, Campanulaceae, Caryophyllaceae) in urbanized and semi-natural habitats in Switzerland, per plant. Subscript numbers denote error term used for calculating F- and p-values. Species without flower visitors were excluded from these analyses.

Error terms: 1: Species; 2: Location; 3: Species × Habitat type; 4: Species × Location; 5: Observation session; 6: Residuals. #: covariables had a positive effect on response variable. d.f.: degrees of freedom, *0.01<p<0.05

Online Resource Table 5Resultsofthestatisticalanalysesof pollinator visitation per plant of native andrelatedintroduced plant species (without Caryophyllaceae) in urbanizedand semi-natural habitats in Switzerland. Subscriptnumbersdenoteerrortermusedforcalculatingquasi-F- andp-values.

Error terms: 1: Species; 2: Location; 3: Species × Habitat type; 4: Species × Location; 5: Observation session; 6: Residuals. #: covariables had a positive effect on response variables. †: The proportion of variation explained by each factor can be estimated using the change in deviance (i.e. the mean deviance multiplied by the d.f.) relative to the total deviance; d.f.: degrees of freedom, ***p<0.001, **0.001<p<0.01, *0.01<p<05.

Online Resource Table 6Results of the statistical analysis of the time a pollinator spent on native and related introduced plant species (without Caryophyllaceae) in urbanized and semi-natural habitats in Switzerland, per plant. Subscriptnumbersdenoteerrortermusedforcalculating F- andp-values. Specieswithoutflowervisitorswereexcludedfromtheseanalyses.

Error terms: 1: Species; 2: Location; 3: Species × Habitat type; 4: Species × Location; 5: Observation session; 6: Residuals. #: covariables had a positive effect on response variable. d.f.: degrees of freedom, *0.01<p<0.05

Online Resource Test for differences in pollinator-community composition

To test whether the composition of the pollinator community differed among native, invasive alien and non-invasive alien species, and between urbanized and semi-natural habitat types, we used non-metric multidimensional scaling (NMDS) and permutational multivariate analysis of variance (PERMANOVA), as implemented in the vegan package (Oksanen et al. 2011) of the statistical software R (R Development Core Team 2009). For both the NMDS analysis (Vegan function metaMDS) and the PERMANOVA (Vegan function adonis), we first calculated a Gower dissimilarity matrix based on a matrix containing the number of visits by each group of pollinators (honeybees, bumblebees, other bees, hover flies, flies, wasps, other dipterans, butterflies, beetles, and other pollinators) to each of the 252 individuals plants in our study. For the NMDS analysis, we set the number of axes equal to three, as increasing the number of axes did not greatly reduce the stress value, and we set the maximum number of random starts equal to 50. The final stress value was 0.142. To assess whether the NMDS values differed consistently among the three groups of species and between the two habitat types, we used linear models. For the PERMANOVA, the P-values were based on 999 permutations. Because the three groups of species and the two habitat types were repeated for each of the plant families, data permutations were done within families. In this analysis, it was not possible to include interactions of family with the other factors.

Visual inspection of the NMDS plots (Appendix Figure 1) reveals that there was more spread in the NMDS values among species when the observations were done in urbanized (open symbols) instead of semi-natural habitats (closed symbols). Similarly, at least in the plot of NMDS2 vs NMDS3, the alien species (triangles) showed more spread in NMDS values than the native species (circles). Although the analyses of the NMDS values revealed significant effects of the number of flower unites on NMDS values, variation among species and variation among locations, these analyses did not reveal significant differences among the native, invasive alien and non-invasive alien species, and also not between the urbanized and semi-natural habitat types (Appendix Table 6).

The PERMANOVA, however, revealed significant differences in pollinator-group composition between native and alien species, irrespective of whether the latter are invasive or non-invasive, and between the urbanized and semi-natural habitat types (Appendix Table 7). Inspection of the NMDS plots (Appendix Figure 1) suggest that these differences in pollinator-community composition mainly reflect higher frequencies of other bees, hover flies, flies and beetles on native than on alien species, and in semi-natural than in urbanized habitats.

References

Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, O’Hara RB, Simpson GL, Solymos P, Stevens MHH, Wagner H (2011) Vegan: Community Ecology Package. R package version 2.0-0.

R Development Core Team (2009) R: A language and environment for statistical computing.R Foundation for Statistical Computing, Vienna, Austria.

Online Resource Figure 2The three non-metric multidimensional scaling (NMDS) axes derived from a Gower-dissimilarity matrix of differences in composition of the pollinator groups visiting the plants in our study. The NMDS analysis was based on a pollinator-visitation matrix including all 252 individual plants in our study. However, for clarity we only present average NMDS values per status group (circles: native species, upward triangles: invasive alien species, downward triangles: non-invasive alien species) for each of the five plant families (black: Asteraceae, blue: Campanulaceae, green: Malvaceae, pink: Caryophyllaceae, red: Balsaminaceae) separately for the two habitat types (open symbols: urbanized habitats, closed symbols: semi-natural habitats). We also plotted the positions of the different pollinator groups in the plots.

Online Resource Table 7Results of the statistical analyses of NMDS (non-metric multidimensional scaling) values of individual plants of related native and alien plant species in urbanized and semi-natural habitats in Switzerland. Subscript numbers denote error term used for calculating F- and p-values.

Error terms: 1: Species; 2: Location; 3: Species × Habitat type; 4: Species × Location; 5: Observation session; 6: Residuals. d.f.: degrees of freedom, **0.001<p<0.01, *0.01<p<0.05

Online Resource Table 8Results of a PERMANOVA (permutational multivariate analysis of variance) on the matrix of plant individuals and pollinator groups to test for differences in composition of the pollinator-group community on native, invasive alien and non-invasive alien species in urbanized and semi-natural habitats. Interactions of family with other factors could not be included in the model. P-values are based on 999 permutations.

***p<0.001, **0.001<p<0.01, *0.01<p<0.05